Cancer precursors signs are the so called pre-cancerous states, which are curable if they are detected at an early stage. In the opposite case the lesion can progress in depth, resulting in the development of invasive cancer and metastases. At this stage, the possibilities of successful therapy are dramatically diminished. Consequently, the early detection and the objective identification of the severity (stage) of the precancerous lesion are of crucial importance.
The conventional clinical process of optical examination have very limited capabilities in detecting cancerous and pre-cancerous tissue lesions. This is due to the fact that the structural and metabolic changes, which take place during the development of the disease, do not significantly and with specificity alter the color characteristics of the pathological tissue.
In order to obtain more accurate diagnosis, biopsy samples are obtained from suspicious areas, which are submitted for histological examination. However, biopsy sampling poses several problems, such as: a) risk for sampling errors associated with the visual limitations in detecting and localizing suspicious areas; b) biopsy can alter the natural history of the intraepithelial lesion; c) mapping and monitoring of the lesion require multiple tissue sampling, which is subjected to several risks and limitations; d) the diagnostic procedure performed with biopsy sampling and histologic evaluation is qualitative, subjective, time consuming, costly and labor intensive. In recent years there have been developed and presented quite a few new methods and systems in an effort to overcome the disadvantages of the conventional diagnostic procedures. These methods can be classified in two categories: a) Methods which are based on the spectral analysis of tissues in vivo, in an attempt to improve the diagnostic information b) Methods which are based on the chemical excitation of tissues with the aid of special agents, which have the property to interact with pathologic tissue and to alter its optical characteristics selectively, thus enhancing the contrast between lesion and healthy tissue.
In the first case, the experimental use of spectroscopic techniques as a potential solutions to existing diagnostic problems, is motivated by their capability to detect alterations in the biochemical and/or the structural characteristics, which take place in the tissue during the development of the disease. In particular, fluorescence spectroscopy has been extensively used in various tissues, where the later are optically excited with the aid of a light source (usually laser), of short wave length (blue-ultraviolet range) and their response is measured as fluorescence intensity vs. wavelength.
Garfield and Glassman in U.S. Pat. No. 5,450,857 and Ramanajum et al in U.S. Pat. No. 5,421,339 have presented a method based on the use of fluorescence spectroscopy for the diagnosis of cancerous and pre-cancerous lesions of cervix. The main disadvantage of fluorescence spectroscopy is that the existing biochemical modifications associated with the progress of the disease are not manifested in a direct way as modifications in the measured fluorescence spectra. The fluorescence spectra contain limited diagnostic information of two basic reasons: a) Tissues contain non-fluorescent chromophores, such as hemoglobin. Absorption by such chromophores of the emitted light from fluorophores can result in artificial dips and peaks in the fluorescence spectra. In other words the spectra carry convoluted information for several components and therefore it is difficult to assess alterations in tissue features of diagnostic importance; and b) The spectra are broad due to the fact that a large number of tissue components are optically excited and contribute to the captured optical signal. As a result the spectra do not carry specific information for the pathologic alterations and thus they are of limited diagnostic value.
The latter is expressed as low sensitivity and specificity in the detection and classification of tissue lesions.
Aiming to enhance the sensitivity and specificity of the captured information, Ramanujan et al in the Patent No. WO 98/24369 have presented a method based on the use of neural networks for the analysis of the spectral data. This method is based on the training of a computing system with a large number of spectral patterns, which have been taken from normal and from pathologic tissues. The spectrum that is captured each time is compared with the stored spectral data, facilitating this way, the identification of the tissue pathology.
R. R. Kortun et al, in U.S. Pat. No. 5,697,373, seeking to improve the captured diagnostic information, have presented a method based on the combination of fluorescence spectroscopy and Raman scattering. The last has the capability of providing more analytical information, it requires however complex instrumentation and ideal experimental conditions, which substantially hinder their clinical use. It is generally known that tissues are characterized by the lack of spatial homogeneity and consequently the spectral analysis of distributed spatial points is insufficient for the characterization of their status.
Dombrowski in U.S. Pat. No. 5,424,543, describes a multi-wavelength, imaging system, capable of capturing tissue images in several spectral bands. With the aid of such a system it is possible in general to map characteristics of diagnostic importance based on their particular spectral characteristics. However, due to the insignificance of the spectral differences between normal and pathologic tissue, which is in general the case, inspection in narrow spectral bands does not allow the highlighting of these characteristics and even more so, the identification and staging of the pathologic area.
D. R. Sandison et al, in U.S. Pat. No. 5,920,399 describe an imaging system, developed for the in vivo investigation of cells, which combines multi-band imaging and light excitation of the tissue. The system also employs a dual fiber optic bundle for the transferring of the emitted from the source light onto the tissue and the remitted light from the tissue to the optical detector. These bundles are placed in contact with the tissue, and various wavelengths of excitation and imaging are combined in attempt to enhance the spectral differentiation between normal and pathologic tissue.
In U.S. Pat. No. 5,921,926, J. R. Delfyett et al have presented a method for the diagnosis of diseases of the cervix, which is based on the combination of Spectral Interferometry and Optical Coherence Tomography (OCT). This system combines three-dimensional imaging and spectral analysis of the tissue.
Moreover, several improved versions of colposcopes have been presented, (D. R. Craine et al, U.S. Pat. No. 5,791,346 and K. L. Blaiz U.S. Pat. No. 5,989,184) in most of which, electronic imaging systems have been integrated for image capturing, analysis of tissue images, including the quantitative assessment of lesion's size. For the enhancement of the optical differentiation between normal and pathologic tissue, special agents are used in various fields of biomedical diagnostics, which are administered topically or systematically. Such agents are acetic acid solution, toluidine blue, various photosensitizers (porphyrines) (S. Anderson Engels, C. Klinteberg, K. Svanberg, S. Svanberg, In vivo fluorescence imaging for tissue diagnostics, Phys Med. Biol. 42 (1997) 815-24). The provoked selective staining of the pathologic tissue is owed to the property of these agents to interact with the altered metabolic and structural characteristics of the pathologic area. This interaction enhances progressively and reversibly the differences in the spectral characteristics of reflection and/or fluorescence between normal and pathologic tissue. Despite the fact that the provoked selective staining of the pathologic tissue is a dynamic phenomenon, in clinical practice the intensity and the extent of the staining are assessed qualitatively and statically. Furthermore, in several cases of early pathologic conditions, the phenomenon of temporary staining after administering the agent, is short-lasting and thus the examiner is not able to detect the provoked alterations and even more so, to assess their intensity and extent. In other cases, the staining of the tissue progresses very slowly, with the consequence of patient discomfort and creation of problems for the examiner in assessing the intensity and extent of the alterations, since they are continuously changing. The above have as direct consequence, the downgrading of the diagnostic value of these diagnostic procedures and thus its usefulness is limited to facilitate the localization of suspected areas for obtaining biopsy samples.
Summarizing the above mentioned, the following conclusions are drawn:
a) Various conventional light dispersion spectroscopic techniques (fluorescence, elastic, non elastic scattering, etc) have been proposed and experimentally used for the in vivo detection of alterations in the structural characteristics of pathologic tissue. The main disadvantage of these techniques is that they provide point information, which is inadequate for the analysis of the spatially non-homogenous tissue. Multi-band imaging has the potential to solve this problem, by providing spectral information (of lesser resolution as a rule) but in any spatial point of the area under examination. These techniques (imaging and non-imaging) however, provide information of limited diagnostic value, due to the fact that the structural tissue alterations, which are accompanying the development of the disease, are not manifested as significant and characteristic alterations on the measured spectra. Consequently, the captured spectral information cannot be directly correlated with the tissue pathology, a fact which limits the clinical usefulness of these techniques.b) The conventional (non-spectral) imaging techniques provide the capability of mapping characteristics of diagnostic importance in two or three dimensions. They are basically used for measuring morphological characteristics and as clinical documentation tools.c) The diagnostic methods which are based on the selective staining of pathologic tissue with special agents allows the enhancement of the optical contrast between normal and pathologic tissue. Nevertheless they provide limited information for the in vivo identification and staging of the disease.
Given the fact that the selective interaction of pathologic tissue with the agents, which enhance its optical contrast with healthy tissue is a dynamic phenomenon, it is reasonable to suggest that the capture and analysis of the characteristics of this phenomenon's kinetics, could provide important information for the in vivo detection, identification and staging of tissue lesions. In a previous publication by the inventors (C. Balas, A. Dimoka, E. Orfanoudaki, E. Koumandakis, “In vivo assessment of acetic acid-cervical tissue interaction using quantitative imaging of back-scattered light: Its potential use for the in vivo cervical cancer detection grading and mapping”, SPIE-Optical Biopsies and Microscopic Techniques, Vol. 3568 pp. 31-37, (1998)), measurements of the alterations in the characteristics of the back-scattered light as a function of wave-length and time are presented. These alterations are provoked in the cervix by the topical administration of acetic acid solution. In this particular case, there was used as an experimental apparatus, a general-purpose multi-spectral imaging system built around a tunable liquid crystal monochromator for measuring the variations in intensity of the back-scattered light as a function of time and wavelength in selected spatial points. It was found that the lineshapes of curves of intensity of back-scattered light versus time, provide advanced information for the direct identification and staging of tissue neoplasias. Unpublished results of the same research team support that similar results can also be obtained with other agents, which have the property of enhancing the optical contrast between normal and pathologic tissue. Nevertheless, the experimental method employed in the published paper is characterized by quite a few disadvantages, such as: The imaging monochromator requires time for changing the imaging wavelength and as a consequence it is inappropriate for multispectral imaging and analysis of dynamic phenomena. It does not constitute a method for the mapping of the grade of the tissue lesions, as the presented curves illustrate the temporal alterations of intensity of the back-scattered light in selected points. The lack of data modeling and parametric analysis of the characteristics of the phenomenon kinetics in any spatial point of the area of interest restrict the usefulness of the method in experimental studies and hinder its clinical implementation. The optics used for the imaging of the area of interest are of general purpose and do not comply with the special technical requirements for the clinical implementation of the method. Clinical implementation of the presented system is also hindered by the fact that it does not integrate appropriate means for ensuring the stability of the relative position between the tissue surface and image capturing module, during the snapshot imaging procedure. This is very important since small movements of the patient (i.e. breathing) are always present during the examination procedure. If micro-movements are taking place during successive capturing of images, after application of the agent, then the spatial features of the captured images are not coincide. This reduces substantially the precision in the calculation of the curves in any spatial point, that express the kinetics of marker-tissue interaction.